Hello everyone
we are using the following scheme for DHCP relaying:
VLANs have two relay-agents (one on every Coreswitch)
External RED networks have the UTM as relay agent
External UTMs have switches in their LANs with relay agents
All of them relay to a DHCP failover cluster in one way or another (see attached schematic)
What I now see wiresharking around on my DHCP servers is relays from the UTMs IP within the DHCP-Server subnet for bootp.hw.mac_addr entries that are local clients
So it seems as if the UTM is re-relaying stuff the cores already relayed
Doesn't the UTM check for the relay-agent header?
If I understand the config dialog of the UTM right you set up the following:
1) Server(s) to relay to
2) Interfaces on which the UTM will listen for DHCP broadcasts
The later causes that both the RED interfaces as well as an interface of the UTM in the DHCP-server subnet need to be configured
Am I correct on this?
Does my config make sense at all or is there a better way to do it?
I it intended that the UTM re-relays or is this a bug?
Best regards
Edit: as expected leaving the path of rightousness is never a good idea. I set the DHCP relay on the core switches back to 2 relay agents per VLAN directly relaying to the DHCP-server nodes instead of the networks broadcast address.
So that is now standard conform. Nevertheless the other questions remain
This thread was automatically locked due to age.